Position Sizing Is a Different Question From Stock Selection

Every investor who has ever owned a stock that turned into a multibagger ends up asking the same question afterward: why didn't I buy more of that one. It happens to retail traders and seasoned HNIs alike, without exception. The mind runs the counterfactual, imagines the portfolio if that one position had been twice or three times the size, and then measures the actual outcome against that imagined, better version of the past.

That comparison feels like analysis. It's actually just hindsight doing what hindsight always does, rewriting the past into its best possible version and then judging the present against a standard that was never knowable in advance. Position sizing has to be decided before the outcome exists, not after it's already known which stock was going to be the big winner.

Two different questions, not one

Everything covered in this series so far, the fundamental filter, the technical trigger, relative strength, the multi-year base, answers exactly one question: should this stock be owned at all. That's stock selection.

Position sizing asks something entirely different: how much of the portfolio should this one idea be allowed to occupy. A great answer to the first question says nothing at all about the second one, and treating them as the same decision is exactly how a good stock pick turns into a damaging portfolio outcome.

Conviction is not a risk measurement

However convinced anyone is about a particular stock, no single position should ever be allowed to grow into a majority of the portfolio, more than half of total capital in one name is a hard line worth never crossing, regardless of how strong the case looks.

The reason is straightforward. Even a thoroughly researched, genuinely well-selected stock can fail for reasons that have nothing to do with the quality of the analysis behind it, a lawsuit, a regulatory action, a fraud that wasn't visible from the outside, a sudden macro shock. Conviction measures how much an investor believes in an idea. It says nothing about how much of a portfolio should be exposed to the possibility of being wrong about it, and those are two different quantities that shouldn't be allowed to blur into each other.

So how does position size actually get decided

Three things drive the decision: whether the portfolio is equal weighted or variable weighted, and how many stocks it holds at all.

An equal weighted portfolio assigns the same fixed position size to every stock it holds, regardless of how strongly any individual idea is favoured. Run a portfolio of ten stocks this way, and the math does the risk management automatically. A single stock that doubles moves the total portfolio up by 10%. A single stock that hits a 10% stop loss moves the total portfolio down by only 1%. No individual name, however good or bad it turns out to be, can single-handedly make or break the outcome. That's not an accident, it's the entire point of equal weighting.

The number of stocks held matters just as much as how they're weighted. Too few names, and each position becomes so large that a single mistake can do real damage to the whole portfolio, defeating the purpose of holding more than one stock in the first place. Too many names creates a different, quieter problem, one that shows up less in the math and more in the ability to actually manage what's owned. A doctor once described his own portfolio as holding seventy stocks or more, essentially every well known name that had ever come up in conversation. The trouble wasn't the number itself so much as what it revealed: most of those positions were there because someone had recommended them at some point, not because he had studied them or held any real conviction in why each one belonged in the portfolio. Owning that many names makes it practically impossible to actually track or analyse each one properly, and a portfolio built that way ends up performing however it happens to perform, with no real understanding of why. A reasonable range sits somewhere between roughly four or five stocks at the low end and fifteen to eighteen at the high end, wide enough to diversify away single-stock risk, narrow enough that every position can still be genuinely known and watched.

Variable weighting reintroduces the very trap this is meant to solve

Some investors prefer to size positions based on conviction, allocating more to ideas backed by deeper study and less to others, rather than weighting everything equally. This isn't inherently wrong, but it quietly reopens the exact psychological trap this entire discussion started with. Conviction is a feeling, and feelings have a well documented habit of growing after the fact, exactly the way the multibagger regret does. Anyone using variable weighting needs a hard ceiling on it, the same 50% rule and reasonable stock-count range apply regardless, rather than letting position size drift upward purely because an idea feels increasingly right.